With Great Power...

[Shamus Young]

With Great Power...

I said last week [http://www.escapistmagazine.com/articles/view/columns/experienced-points/10904-The-Indies-Will-Ruin-Everything] that this new console generation threatens to bust the already tenuous budget problems that developers are having. But this doesn't mean that more processing power is a bad thing.

Read Full Article

 

[BrotherRool]

I'm pinning my hopes on that last one. In particular I'm hoping the design savings might be enough to enable some more midbudget games to exist again and to allow larger indie's to push what they want to do

 

[Dr.Awkward]

Question here: Has anyone researched the resources required to render numerous vector-based textures in a 3D space? It'd solve the "texture" problem as they can scale up and down with little to no loss in detail.

 

[Rad Party God]

People should use more Source-style engines.

Hear me out, I like pretty graphics as much as the next person, but just as Shamus pointed out, we don't need to squeeze every last byte the new hardware offers, we can perfectly use current (last?) gen assets and increase them!, imagine a game with tons of characters/objects/explosions on-screen in 1080p and running at 60 fps, with little to no loading times...

That's the kind of experience I'd like to have, instead of waiting for 30 minutes for the game to initially install itself, running at 30 fps, shooting in linear corridors and lasts for 6 hours. Wouldn't you prefer to compromise a little graphical fidelity for a quality gaming experience?.

That's what got me so mad when Capcom said game development has become more expensive with the new hardware. [small]Seriously, fucking idiots.[/small]

 

[mjc0961]

In GTA we could have streets packed with traffic and sidewalks packed with pedestrians.

That sounds like a really cool idea. Somebody just needs to remind Rockstar how to actually complete and ship a game properly again so they can get some consumer trust back. If GTA VI, whenever it comes out, isn't a fully completed game with all major bugs removed on day one (for example, my cars don't go fucking missing from my garage which was a blatantly obvious and easily reproducible bug that never should have existed in a game that got a 6 month delay so they could "polish" it) with a fully functional online mode with all advertised features present if they do another online mode, then I won't be giving them a penny again.

But yeah, maybe that's another idea that might be nice. Do all the tips in the article to get your game done quicker and with less money spent. This will let you focus less on stupid stuff that doesn't matter and more time on shipping a completed product!

 

[UNHchabo]

...imagine a game with tons of characters/objects/explosions on-screen in 1080p and running at 60 fps, with little to no loading times...

Or if we did squeeze the hardware a little bit through max player counts (although I imagine bandwidth is the limiting factor there in most cases). Red Orchestra 2 and Battlefield 4 can handle 64 players, but imagine if they could raise that limit; you could have massive battles on a simply huge amount of land!

 

[Norix596]

One of my favorite ideas I've heard from the podcast was to take new technology to not make, say, Skyrim look better but to keep the level of graphics the same but removing and shortening some loading screens between houses and areas.

 

[Dr.Awkward]

People should use more Source-style engines.

For that to happen, Hammer's problems need to be fixed (the program itself needs major updating), and plugins need to be created for popular software that can access Source's files.

For one, a lot of frustration would be absolved if it was possible to build and modify conditions for trigger primitives through a node graph instead of having to put all the needed entities on the map itself, along with the option of mipmap and pathing nodes being automatically generated (although finally upgrading to navmeshes + helper nodes would be better). Secondly, something as simple as extrusion & subdivision would be well-appreciated by everyone since they can now easily work from a simple primitive to make the commonly irregular shapes more interesting maps utilize.

 

[Saulkar]

A couple points are somewhat moot. I know I am being pedantic but I must point out that making textures higher res (such as for a keyboard) is not at all more time intensive, in fact it is much, much, much easier!

Spoiler: I finished this back in the summer

These textures range in size from 512x512 all the way up to 4096x4096 and the larger ones are always easier because you do not need to screw around with the ungodly artifacts that occur when you strategise a low resolution texture from scratch. It is ultimately easier to do it all highres then scale it down in Photoshop. It is likely that, that keyboard texture was higher resolution but scaled down before being imported into the engine.

Lastly, models from 2008 are just as detailed as the ones today. The only difference is that the 1-20 million polygon projection model gets baked onto a slightly higher resolution base mesh. With more processing power you could bake less of the high resolution mesh (stuff like bolts instead of whole sections of stacked detail) and use it as the base instead of creating a brand new low-poly one and have the engine dynamically decimate it the farther away you get where you cannot see artifacting.

 

[rapidoud]

I agree completely Shamwow, we need more detail. This is the main reason why Supreme Commander excited me so much when I played it. The models weren't even that great looking at the time but you could have so many on-screen that the game was a blast to play. Unfortunately, people complain about pixels too much for any AAA devs to risk it.

 

[Nixou]

How do games like the Crysis and Witcher series (with extreme graphical fidelity and custom engines) still make money

As far as the Witcher is concerned: paying polish salaries and putting the company's head office assets in a tax heaven favored by russian mobsters [http://en.wikipedia.org/wiki/2012?13_Cypriot_financial_crisis]

 

[TiberiusEsuriens]

I'm pinning my hopes on that last one. In particular I'm hoping the design savings might be enough to enable some more midbudget games to exist again and to allow larger indie's to push what they want to do

I agree. "AA" Games are so much better more often than "AAA" blockbuster titles. They are also where the industry saw the most diversity back in the PS2 era. These days we see all "AAA" games are action-masturbation titles, but even the "A" indie scene is back to simply chasing and mimicking the latest fads. Last year it was pixel art games, this year it's Minecraft-like survival games. I mean, some of those are still pretty decent, but it's really exciting when a medium sized studio shows off a completely different game, gives clear reasons why they want to make it, and sticks to the plan to avoid budget and feature creep. /rant

 

[Callate]

I don't think "better AI" always naturally translates to "smarter AI", at least in the sense of "AI that consistently kicks the human player's ass". It's not hard at all to make AI that can always hit a headshot, and I don't think anyone wants to play against that. But people were very impressed with the way AI in F.E.A.R. seemed to do things like topple furniture for cover, fall back from weakened positioning, and flank. AI that can make intelligent assessments of the local terrain- especially without hard-wired foreknowledge of that terrain- could be great. AI that makes human-like mistakes is frequently more impressive than AI that shows machine-like efficiency.

As far as models go, I think someone needs to make a universal "doll" that works like the avatar modeling in a game like Skyrim or Saint's Row, but on a grander scale. If someone could make a model that could convincingly be a hundred different people without anyone recognizing it as the same model with different texturing, I think it could be as big an advance as Havok physics. But it would need more than variable textures, higher brow-ridges, or a fatter midsection- you would need to create a model that could have visibly different weight distribution that would effect how it moved, different centers of gravity, different torso-to-leg ratios. If you could make one model that could be Lara Croft, Andre the Giant, or Gollum by virtue of a few minutes of tweaking, I think you could streamline a significant amount of character design, at least for NPCs.

 

[XMark]

I'd like to see more games switch to the "everything is one gigantic seamless world" approach, which will definitely be easier to do this generation than the last one. Dead Space 2 and 3 and Dark Souls are both great examples on how this was accomplished last generation. This time around, we won't need as many hidden "loading elevators".

 

[RandV80]

One of my favorite ideas I've heard from the podcast was to take new technology to not make, say, Skyrim look better but to keep the level of graphics the same but removing and shortening some loading screens between houses and areas.

Yes Ultima IX was able to do this like 15 years ago, with the new gen it's time for Bethesda to drop the load screen when walking into a simple house.

I also like the idea of filling the worlds with more people. Even if they're just milling about it creates better ambiance. It was on the dated Wii hardware but The Last Story is a good example of this, you still wouldn't exactly call it crowded but compared to other games the city had plenty of people simply wandering about. I actually haven't played any but I guess the Assassins Creed series does this as well.

I guess you have to ask yourself what's better. An TES:Oblivion style game where every citizen has a purpose and a routine they follow throughout the day, but you only have the processing power/time to put 30 of them in a town, or making less AI intensive/interesting people but filling that same town with 300 of them? Or better yet, since we're talking new gen now combine the two.